Application of Iteratively Truncated Singular Value Decomposition to Computation of Teleseismic Receiver Functions

Receiver function analysis is perhaps the most popular tool amongst seismologists for estimating the receiver-side crustal structure using three-component recordings of teleseismic P-wave coda. Several methods have been suggested for successful construction of receiver function time series, and almost all of them involve deconvolution, the inverse of the convolution operation, in either the time or frequency domain. This paper presents an automated time-domain deconvolution approach employing truncated singular value decomposition iteratively during the computations. This theoretically straightforward approach does not need any regularization parameter, and results from its application to noise-containing synthetic seismograms demonstrate its effectiveness. A high level of compatibility between the receiver functions calculated using the recordings of the Lodumlu (LOD) station, located in central Anatolia, and results from previous seismological studies provides further evidence for the functionality of this new algorithm. Based on hypothetical experiments and real data applications, iteratively truncated singular value decomposition may be considered a reliable alternative to conventional receiver function calculation schemes. Simplicity, stability, and the need for minimal human interaction are the main advantages of the proposed technique.